Skip to content

makhembu/watu

Repository files navigation

Watu — AI CV Screening

Analyze how well a candidate's CV matches a job description. Watu (Swahili for "people") uses AI to score fit, identify strengths and gaps, and suggest interview questions.

Quick Start

git clone https://github.com/makhembu/watu
cd watu
cp .env.example .env
# Add your Gemini API key to .env
npm install
npm run dev
# Open http://localhost:3000

How It Works

  1. Paste a job description and a CV
  2. AI analyzes both for skill overlap, experience fit, and gaps
  3. Returns a score (1-10), strengths, gaps, recommendations, and interview questions
  4. Works without an API key (rule-based fallback)

Features

  • AI-powered analysis using Google Gemini
  • Rule-based fallback — works even without an API key
  • Dark mode UI — easy on the eyes
  • No data storage — privacy by design
  • Rate limited — 10,000 character input limit
  • Export-ready — scores and analysis you can screenshot

Stack

  • Next.js 15 (App Router)
  • TypeScript
  • Google Gemini API
  • Tailwind CSS 4
  • Server-side API route

Deployment

# Deploy to Vercel
npx vercel --prod

Set GEMINI_API_KEY in your Vercel environment variables.

Why

Hiring in Africa often relies on manual CV screening. Watu makes the first pass instant, consistent, and bias-reduced. Built for recruiters, HR teams, and hiring managers who want a second opinion.

Roadmap

  • PDF upload support
  • Batch CV comparison
  • Export reports (PDF)
  • API key management UI
  • Multiple AI model support

About

AI-powered CV screening tool — score candidate-job fit, identify gaps, suggest improvements

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages